Approach to cartilage injury in the anterior cruciate ligament-deficient knee

Articular cartilage: function 

Successful human athletic performance requires the optimal function of our articulations. Proper joint function requires not only adequate strength and stability, but a smooth, gliding articular surface to allow an effortless range of motion. Dysfunction of this natural bearing surface results in pain, swelling, and limitation of function [6]. It has also been known that injury to this structure may result in progressive degeneration into osteoarthritis [7].

Articular cartilage: structure 

The junction of calcified and noncalcified tissues has increasingly become implicated as the site of the essential lesion in articular cartilage injury. The limited ability of articular cartilage to repair itself has come under increased scientific scrutiny. Cartilage science has blossomed over the past 10 years, resulting in a dramatic advancement of our understanding of the mechanics and anatomy of cartilage injury. We now have a better understanding of articular cartilage as a composite tissue with specific mechanical and physiologic properties.

Articular cartilage consists of a highly specialized extracellular matrix surrounding a sparse population of chondrocytes. Chondrocytes are derived from mesenchymal cells, and it has been demonstrated that perivascular mesenchymal cells in cell culture can be recruited to form chondrocytes or osteocytes under the influence of certain morphogenetic factors [8]. The chondrocytes are responsible for the production and maintenance of the extracellular matrix. Cartilage is aneural and avascular, leaving chondrocyte viability dependent on diffusion of metabolites from the synovial fluid.

The extracellular matrix in articular cartilage is responsible for its smooth gliding properties. The extracellular matrix consists primarily of type 2 collagen, proteoglycans, and water. Type 2 collagen, which makes up 90–95% of the collagen present in articular cartilage, is largely responsible for the ability of cartilage to resist shear stress [9], [10].

The proteoglycans present in cartilage consist of glycosoaminoglycans, including chondroitin sulfate and keratin sulfate side chains bound to a protein core to form the aggrecan macromolecule. The protein core is bound noncovalently to hyaluronate via link protein to form large aggregates of aggrecan-hyaluronate complexes. Glycosaminoglycans in cartilage contain both carboxy-terminal and sulfated ends, which are ionized in solution. These ionized groups result in aggrecan molecules being highly hydrophilic and account for the swelling pressure of cartilage. As a pressurized proteoglycan gel with collagen fiber reenforcement, cartilage is highly resistant to compressive loads.

Articular cartilage is organized into four distinct layers: superficial, middle, deep, and the calcified cartilage [11], [12]. In the superficial layer, cells and collagen are arranged in a plane parallel to the joint surface and proteoglycans are at their lowest concentration. This surface is especially well suited to facilitate the gliding motion of joints and the large shear forces acting at the articular surface. The middle layer has more rounded cells and an oblique orientation to the collagen fibers. In the deepest layer, the cell population is at its largest concentration, as is the proteoglycan concentration. The collagen fibers in the deepest layer are oriented perpendicular to the joint surface. Deepest and adjacent to bone is the calcified cartilage. The calcified cartilage is separated from the noncalcified deep cartilage layer by an eosinophilic staining wavy line referred to as the tidemark. This junction of calcified from noncalcified tissue has been shown in recent cadaver, animal, and mathematical models to be a likely source for the initial injury in cartilage injuries caused by shear stress [13].

Water molecules are delivered on the surface of the articular cartilage when a load is applied. The water is “squeezed” through pores present in the articular cartilage. These water molecules add to the already present synovial-fluid surface film. The fluid-phase portion of cartilage thus adds to the low friction surface layer between the two articular surfaces. Cartilage demonstrates time-dependent deformation under a constant load: the viscoelastic property of creep. It is most likely the result of the water molecules being slowly squeezed out of the proteoglycan gel. Water molecules experience drag, however, as they are squeezed through the porous collagen fiber network toward the articular surface, thereby resulting in cartilage's excellent ability to resist compression. Under normal conditions, cartilage functions smoothly for decades.

Mechanism of injury 

The majority of ACL tears occur through nonimpact rotational deceleration mechanisms.. At the time of ACL failure, it has been shown that the tibia can sublux anteriorly and impact the lateral femoral condyle. Thus, initial ACL disruption also results in high shear forces across the tibiofemoral articulation. Additionally, the subsequent instability following ACL disruption can allow for increased loads to be transmitted to articular surfaces.

Analytic studies by Mow et al have demonstrated that articular cartilage behaves as a biphasic material. Consequently, shear and blunt forces are manifested at the junction of the uncalcified and calcified cartilage [13]. Utilizing this mathematical model, the authors were unable to derive a situation whereby shearing forces would produce surface failure (abrasion) without first causing damage at the tidemark.

This has been demonstrated in a cadaver model in which loads with variable speed and force were applied to femoral condyles [14], [15]. The low-speed, low-energy injuries resulted in injuries below the articular surface; at the junction of the calcified and noncalcified tissues. A zone of shear forces radiates from the impact site, producing injury to the junction of calcified and noncalcified tissues.

Animal studies have supported this concept. Where low-energy blows produce selective injury to the deep structures [14], [15], [16], they leave the articular surface intact. This data implies that the junction between calcified and noncalcified tissues may represent the “weak link” in articular cartilage's ability to resist shear forces. Transarticular loading of canine metacarpal-phalangeal joints has resulted in cracks appearing in the calcified cartilage [17] that went on to osteoarthritis, thereby strengthening the argument that injury to the deeper layers may ultimately lead to articular degeneration. Recently, this phenomenon has been demonstrated in athletes [18] and in patients with a history of mechanical symptoms [6], [19], [20]. Many of these patients demonstrate arthroscopic evidence of lesions that appeared to have a significantly larger deep component than the immediately evident surface lesion. On probing, the uncalcified tissue appeared to have “peeled” off from the calcified layers.

The correlation between an acute shear-force injury, such as during disruption of the ACL cartilage lesions, is not definite. There is increasing evidence, however, of the poor prognosis of predominant shear-force injuries [16], [18]. Indeed, a reconstructed ligament may restore stability to the knee, but the prognosis of the articular surface may depend on factors other than stability [1]. Though an unstable knee is functionally limiting, it can also lead to subsequent meniscal injury. Meniscal pathology resulting in meniscectomy is known to correlate with joint degeneration, but initially stable or stable reconstructed knees with intact menisci have also progressed to joint degeneration [1]. The origin of this pathology may be the initial, silent insult to the articular cartilage at the time of ACL disruption. Additionally, athletes involved in pivoting sports have developed cartilage lesions without ligament or meniscal pathology [18]. The high shear forces at the level of the tidemark in aggressive pivoting athletes may be sufficient in some situations to induce injury to the cartilage.

Progression of occult articular cartilage injuries is an area of active scientific investigation in which much is yet to be learned. One possible explanation focuses on the potential space created between the calcified and noncalcified cartilage at the time of injury. As the joint is loaded, water molecules may be delivered to this potential space as the cartilage undergoes viscoelastic creep [13]. Thus pressurized, the potential space may expand at the level of the tidemark with each loading of the joint. Eventually, the enlarged potential space may communicate with the joint surface, and a surface lesion will become evident. When probed, this small lesion will appear to have delaminated from the calcified tissue, revealing a much more extensive deep lesion.

Another theory postulates that the biologic response to injury may result in a construct more poorly suited to tolerate physiologic loads. A well-designed animal study found that impulse loading of rabbit knees resulted in subchondral plate fractures that healed into a mechanically stiffer construct (via tidemark advancement) than the normal joint [21]. These joints went on to joint degeneration with time.

Though abrasive and third-body wear does not appear to cause chondral lesions, it will certainly result in progression of articular injury and degeneration, and, eventually osteoarthritis. Acute injury, progression of injury, multiple small insults, absent menisci, alignment, and anatomy are among the multitude of variables that may eventually result in osteoarthritis. An articular lesion that progresses or does not heal could, however, certainly result in a roughened articular surface that could abrade the surface with which it articulates.

Thus, disruption of the ACL can either be associated with acute chondral injury or can result in repetitive abnormal joint forces that subsequently produce additional chondral injury. There may be damage at the tidemark, furthermore, that may not be evident at initial clinical evaluation. This may lead to cartilage deterioration despite stabilization. It is important, however, to realize that isolated chondral lesions are often reported patients who engage in deceleration maneuvers, and some chondral lesions found at ACL reconstruction may have been preexisting and asymptomatic.

Healing 

Injury to the cartilage at either the surface or the deep layers would not be significant if cartilage had a strong ability to heal. Unfortunately, cartilage has a limited ability for self-repair [19], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31]. Numerous studies have demonstrated that acutely injured chondrocytes respond with the production of collagen and proteoglycans. This increased matrix production, however, is insufficient to fill the injured region and does not result in hyaline cartilage. Furthermore, chondrocytes lack the ability to migrate into the adjacent lesion and produce minimal mitotic activity after adolescence [32].

The primary factor related to cartilage's poor healing response is the lack of blood supply to the chondrocytes. This precludes the formation of fibrin clot and inhibits the inflammatory cascade that most tissues utilize in healing [32].

Bone bruises 

The association of bone bruises with acute ACL injury is well recognized. Bone bruises have been demonstrated by magnetic resonance (MR) imaging in more than 80% of knees after acute ACL rupture [33], [34], [35], [36]. These injuries typically occur in the lateral compartment of the knee on the lateral femoral condyle and posterolateral aspect of the tibia. These impaction injuries necessarily also involve injury to the overlying cartilage. Biopsy of the adjacent cartilage has demonstrated chondrocyte degeneration and loss of proteoglycans [37]. The significance of bone bruises and their relationship to chondral defects is not well understood. An association seems to exist between severe bone bruises and chondral injury. Vellet et al described two types of bone bruises, based on degree of bony involvement, as seen on MR imaging. Reticular lesions, which occurred in 70% of patients, are less severe and appear as hemorrhage and edema in the medullary bone but do not involve damage to subchondral bone. Geographic lesions, occurring in 25% of patients, are more severe and involve signal change contiguous with subchondral bone. The authors found that the reticular lesions tended to resolve after 6–12 months, whereas 62% of the geographic lesions led to “osteochondral sequelae” [17]. Spindler et al found that when MR imaging demonstrated impaction injury to the subchondral bone with signal change in the overlying cartilage layer, there was an increased incidence of arthroscopically apparent cartilage injury with bone bruises on the lateral femoral condyle [35]. Careful evaluation with arthroscopic probing of this region is thus warranted at the time of arthroscopy. An association between bone bruises induced in rabbits and late degenerative change has been demonstrated. Although this has not yet been shown in humans, it should be taken into account in return-to-activity decisions following ACL reconstruction.

Diagnosis 

The presence of a chondral lesion may or may not be known prior to performing an ACL reconstruction. Physical examination in an unstable knee is not reliable in detecting chondral lesions. MR imaging has been shown to have a sensitivity as low as 21% for detecting chondral lesions [18]. Joint effusion [38] and the use of intra-articular gadolinium [39] have been shown to increase the rate of detection, but MR imaging, nevertheless, remains an unreliable diagnostic tool in identifying these lesions. Though new MR imaging techniques (3D-SPGR) have reported increased resolution of cartilage, sensitivity for detecting chondral lesions remains poor (62%, dependent on lesion location and grade), and their availability to most clinicians limited. In the future, imaging techniques that actually detect biochemical and biomechanical changes and matrix degeneration may increase sensitivity [40], [41]. Because the current gold standard for identifying chondral lesions is the arthroscopic viewing and careful probing of articular surfaces, many of these lesions that coexist with ACL injury are often first diagnosed at the time of the definitive ligament reconstructive procedure. This fact can be used to illustrate the importance of the surgeon's need to be prepared to address a possible chondral lesion each and every time an ACL reconstruction is performed.

Following successful ACL reconstruction, occult chondral injuries that were not apparent at the time of reconstruction may produce persistent symptomatology. Patients with articular cartilage lesions may present with a myriad of confusing signs and symptoms, but the most common complaint is knee pain. The pain is generally intermittent and may or may not have been related to an additional injury. The pain may be reproducible with certain activities or within a specific arc of motion [18]. Mechanical symptoms of popping, locking, and catching may occur if a fragment has broken loose or a flap of delaminated cartilage is present. Physical examination should proceed with thorough inspection of the knee and alignment, palpation of structures, range of motion testing, and evaluation of joint stability. Special attention should be paid to the joint line and femoral condyles because tenderness may be present if a lesion extends far enough anteriorly to permit palpation. Clinically evident joint effusion is present in only about 30% of patients, though swelling is a common complaint. Joint crepitation may be present but in only 25% of patients [18]. Lesions in the patella-femoral joint are more difficult to identify. The patella-femoral grind and quadriceps resistance tests are sometimes helpful. These lesions can easily be mistaken, however, for anterior knee pain syndrome, patellar instability, chronic chondromalacia, or early patella-femoral osteoarthritis [6]. Although history and physical examination can reveal some cartilage lesions, they are sometimes difficult to identify. The key to making the diagnosis is to maintain a high index of suspicion.

Numerous classifications have been published describing the appearance of chondral lesions [6], [20], [42], [43], [44], [45], [46]. It is critical, however, to note that most of these classifications involve not only visualizing the articular surface but also noting the consistency of the cartilage when arthroscopically probed. Essentially, most classification schemes describe lesions, either open or closed. Closed lesions are simply the softening of the cartilage on probing, with or without surface changes. Open lesions are classified according to size and depth of involvement. Full-thickness lesions involve true exposed subchondral bone, whereas partial-thickness lesions may involve only the layers superficial to the calcified cartilage. Larger and deeper lesions obviously have a worse prognosis.

Arthroscopy is invaluable in the evaluation of articular cartilage lesions in that it allows direct visualization and classification of the lesion. Coincident meniscal or ligament pathology can be addressed, and, if amenable, the chondral lesion can be treated at the same operative sitting. If a more extensive lesion is found, accurate assessment of the pathology can be attained, thus permitting the patient and physician to pursue an objective evaluation of the various more-complicated treatment options.

Treatment 

Radiofrequency 

The use of radiofrequency (RF) has proliferated recently in orthopedics for various applications in tissue modification such as capsular and ligamentous shrinkage and debridement of meniscal and chondral lesions. RF works by generating ultrasonic waves that produce heat as they pass through the target tissue. This heat results in the denaturing and shrinkage of collagen fibers. Both monopolar RF (MPRF) and bipolar RF (BPRF) devices are commercially available. RF has been used to debride meniscus tears and chondral defects by ablating the tissue flaps and rough edges associated with these lesions. An attractive feature is the ease of accessing hard-to-reach areas with the compact RF probe where mechanical debridement may be difficult. Another feature that makes RF popular is its ability to virtually “melt” or “weld” fragmented tissue to a stable margin. This is theoretically attractive when dealing with certain chondral lesions to avoid the excessive “peeling the orange as mentioned,” that can be associated with mechanical debridement (Fig. 1).

View Large Image Download to PowerPoint

Fig. 1. Thermal coblation of patella chondromalacia with radiofrequency.

The concern about this technique is that it could cause residual thermal damage to the surrounding cartilage and subchondral bone and possibly lead to progression of cartilage degeneration or perhaps even osteonecrosis. Lu reported on the effects of MPRF on partial-thickness cartilage lesions in sheep and found that, though the appearance of the treated surfaces was more smooth when compared with mechanical debridement, scanning electron microscopy revealed significantly more chondrocyte death in the RF group. Studies show that RF induces changes resulting in decreased chondrocyte viability, and progressive loss of proteoglycan over time [48], [49]. Theoretically, BPRF should cause less collateral damage than MPRF because the current flows from electrode to electrode at the probe tip in contrast with monopolar current which travels from probe tip, through the patient, and to the grounding pad. Current studies do not substantiate this, though. Kaplan performed a study, supporting the use of BPRF, with fresh human cartilage from total-knee arthroplasty specimens using hematoxylin and eosin staining and found that the chondrocytes remained in their lacuna when analyzed 6 hours post-trauma [50]. Lu et al, however, did a similar study with confocal laser microscopy and vital cell staining because they felt that light microscopy did not evaluate chondrocyte viability effectively. This study revealed chondrocyte death to a depth of 1.5 mm to 2.5 mm. Depending on the degree of articular thinning in the particular specimen, this sometimes involved penetration all the way down to subchondral bone [51]. Recent forthcoming work by Markel also shows BPRF creating significant chondrocyte death, even to a greater extent than MPRF.

Depth of penetration has been shown to be dependent on temperature, duration, and tissue quality. The collagen denaturation and shrinkage desired for annealing chondral surfaces begins to occur at 65° Celsius, whereas chondrocyte death occurs at only 45°. Exposure times of 5 seconds with MPRF have been shown to create cellular destruction comparable to mechanical debridement, but at least 15 seconds of treatment may be required to smooth the surface adequately, doubling the depth of cellular death [52]. Moller et al has shown that the more extensive the chondromalacia, the greater the vulnerability of the cartilage to thermodestructive effects [53]. This suggests that the undesirable effects of RF may be more pronounced in the very lesions it is used for most. Stein et al have shown that, after 1 year, electocautery chondroplasty actually produces inferior clinical results in grade 3 and 4 chondromalacia when compared with mechanical debridement [54].

We feel that radiofrequency coblation has a limited role in cartilage debridement. It may be appropriate for patellar chondromalacia where partial-thickness fibrillation can be extensive and full-thickness chondrocyte death may be avoided because patellar cartilage is particularly thick. Great caution should be exercised, though, when considering the use of these devices on condylar lesions where cartilage typically is thinner.

Marrow-stimulating techniques 

Several different marrow-stimulating techniques (MST) have been described, and they involve the stimulation of a fibrocartilage repair response by accessing vascular channels through mechanical penetration of the subchondral bone [33], [34], [35], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [53], [54], [55], [56], [57], [58], [59], [60]. The lesion first fills with blood, creating a hematoma and resulting in the production of a fibrin clot [24], [25]. This allows migration of mesenchymal precursor cells. Early on, hyaline-like cartilage with a high proportion of type 2 collagen may be found in the repair matrix [61]. Over time, however, type 1 collagen becomes more prevalent, resulting in a fibrocartilage fill of the lesion [26]. No evidence exists to suggest that the fibrocartilage binds to the surrounding normal hyaline cartilage. It is theorized that the fibrocartilage's beneficial effect is caused by its ability to seal over the lesion, thus diminishing mechanical stress on the subchondral bone. This same sealant effect may also reduce joint effusions by diminishing cartilage debris from floating free into the joint [24].

Techniques of cartilage repair include abrasion, drilling, and “microfracture”. These techniques vary mainly in their method of achieving subchondral penetration. All of these techniques first employ chondral stabilization and debridement as part of the procedure. A significant advantage of these techniques is that they can all be performed arthroscopically on an outpatient basis with minimal morbidity to the patient.

Cartilage lesion abrasion is differentiated from abrasion arthroplasty in that, in chondral lesion abrasion, only the base of an isolated chondral lesion is abraded to penetrate the subchondral bone and stimulate punctate bleeding. Abrasion arthroplasty has been described in arthritic patients as a diffuse deep abrasion of entire joint surfaces to produce a massive bleeding bone surface. By contrast, abrasion of a chondral lesion usually employs a shaver rather than a burr to minimize deep penetration of the subchondral bone. Short-term data (1–2 years) has demonstrated that this technique is effective in athletes in diminishing symptomatology and permitting a return to competitive sports at approximately 10 to 12 weeks [18]. Most athletes did note persistence of dull aching pain after strenuous activity. The overall relief of symptoms in this athletic population was comparable to the preliminary data published on the results of cartilage cell transplant in nonathletes [62].

Another common way of penetrating the subchondral bone at the base of a chondral lesion is through drilling. Theoretically, this technique preserves part of the calcified cartilage layer while creating vascular channels to allow bleeding from the subchondral bone. There are two basic concerns with this technique. The first is that the speed of the drill may generate excessive heat and produce localized osteonecrosis. The second is that the vascular channels themselves result in columns of fibrocartilage that adversely effect later treatment options [63]. In a single report of this technique via open arthrotomy, patients returned to full activity at 6 months with 69% rated as good, 3% fair, and 28% poor [43].

Another alternative technique for cartilage repair involves the use of specially designed “awls” to make perforations in the subchondral bone (Fig. 2). Light arthroscopic shaving to remove debris follows. The theoretical advantage of this microfracture technique is that access to undifferentiated mesenchymal cells is provided without the drawbacks of drilling. Though no clinical studies of this technique have been published, the inventors have presented early animal data on the technique. In horses, microfracture technique generated 50% more type 2 collagen than controls and demonstrated 60% to 90% fill of full-thickness lesions. The inventors have also compared this technique with deep abrasion in horses. Six weeks after treatment, the microfracture group lesions were noted to have a more hybrid cartilage fill that was felt to be more hyaline-like than that of the deep abrasion group [64].

View Large Image Download to PowerPoint View Large Image Download to PowerPoint View Large Image Download to PowerPoint View Large Image Download to PowerPoint

Fig. 2. (A) Chondral lesion.(B) Chondral lesion after debridement and establishment of a stable border. (C) Chondral lesion after microfracture. (D) Fibrocartilage ingrowth one year later.

Favorable short-term results of MST have been reported. Steadman has reported reduction in pain, swelling, and improved function with an average follow-up of 6 years [65]. Longer-term results may not be as satisfactory, however. Fibrocartilage lacks proteoglycan concentration and is thought to be less durable than the more desireable hyaline cartilage [61]. There are no long-term comparative studies on MST but clinical results probably deteriorate over time because of the poor wear characteristics of fibrocartilage. Limited studies on abrasion arthroplasty demonstrate some temporary symptomatic relief but these beneficial effects do not last beyond 2 to 4 years [66], [67]. Examination of failed repairs has revealed soft, fibrillated tissue with central degeneration [68].

MST remains the standard by which other techniques are measured. Advantages of this method include its ease of use, low cost, low morbidity, ability to be performed arthroscopically as an outpatient procedure, and no contraindications with the technique based on lesion size or location. Smaller and relatively acute lesions in the femoral condyle and trochlear areas tend to respond best [69]. Disadvantages to this technique are the prolonged postoperative period of restricted weight bearing and low durability of repair tissue. Biomechanically, fibrocartilage is markedly inferior to normal hyaline cartilage in response to shear and compressive stress [70]. Combined with the stress riser between the fibro and hyaline cartilage, this makes survivability questionable over long periods of time. The predictability of a successful outcome of fibrocartilage producing treatments may be quite variable in different patients.

Autogenous osteochondral grafting 

The rationale behind autogenous osteochondral grafting (AOG) is the vast array of analytical, cadaver, animal, and (recently) human data demonstrating that the common essential lesion in the development of chondral lesions occurs at or near the tidemark where the subchondral plate (subchondral bone and calcified cartilage) interfaces with the uncalcified cartilage. By replacing the entire osteochondral unit, one could theoretically address and remove the actual site of damage that has led to the visible lesion.

The first known report of transplanting articular cartilage bone fragments was by Judet in 1908 for the treatment of osteochondritis descicans [71]. In 1959, Campbell et al reported using fresh osteochondral autografts with little or no destruction occurring over the first year [72]. In 1961, Pap and Krompecher reported the transplantation of articular cartilage fragments and their associated< 5 mm of subchondral bone in 51 dogs. They noted survival of the articular cartilage of up to 2 years provided normal physiologic loads were applied [73]. In 1962, Entin performed autogenous osteochondral transplantation from the foot to the hand and found that the cartilage was not replaced [74].

In 1963, Campbell again reported on the use of osteochondral grafting on 42 dog knees. In this study, grafts were limited to 1 cm wide × 2 cm long, with subchondral bone depth limited to < 5 mm [75]. They noted that the osseous portion healed by 14 days via creeping substitution, and that the articular cartilage appeared grossly and histiologically normal at 2 years postgrafting. They did note, however, that the articular cartilage of the grafts did not bind to the surrounding cartilage. They also speculated that increasing the depth of the subchondral bone might lead to increased necrosis of deep bone.

The healing of osteochondral autografts has been further studied by McDermott et al [12]. It appears that the surface osteocytes in the subchondral bone survive, whereas those deeper than 3 mm are replaced within the first 4 weeks.

Limited biomechanical studies have also been performed to evaluate the effects of graft size on function. It has been shown that a 15% increase in diameter results in a 50% increase in torsional strength of the grafts [76]. The potential unknown effects of radius mismatch must temper this, however.

In 1994, Hangody reported on a technique termed “mosaicplasty” [77]. Mosaicplasty utilizes multiple plugs of autogenous osteochondral graft to fill chondral lesions. This procedure was performed on 122 humans. There were 57 medial femoral condyle lesions, 48 lateral femoral condyle lesions, and 17 patella lesions with > 1 year follow-up. The lesions ranged from 1 to 8 cm2 and grafts varying from 2.7 to 4.5 mm diameter were obtained from the lateral and medial edge of the condyles to fill the defect via an open arthrotomy. At 1 year follow-up, there were no cases of loosening or backing out of the grafts and the average Hospital for Special Surgery (HSS) score was 88.4 (range: 61–100). Second-look arthroscopy was performed in many cases and showed good congruency.

A prospective multicenter study was initiated to follow up 200 patients for 5 years. Lesions treated averaged 143 mm2 (range, 80–25 mm2). A total of 81% were performed arthroscopically, with 19% requiring miniopen technique. So far, patients reported improvement with the technique, and no deterioration of results has been noted at 2-year follow-up [78].

The use of osteochondral autografts has also been described in conjunction with ACL reconstruction. In 29 cases, 1- to 1.5-cm defects of the femoral condyles were filled with osteochondral plugs (5 mm wide × 10–15 mm long) harvested from the intercondylar notch. At 2 to 3 years post-transplantation, the articular cartilage on the transplants appears (via MR imaging and probing) to have survived. The surrounding area is filled with fibrocartilage. Clinically, this has correlated with 19 excellent outcomes in 22 cases [79].

The resurgence in enthusiasm for osteochondral grafting is associated with the increased development of techniques that permit more predictable arthroscopic harvesting and delivery of the plugs. This, however, applies only to limited areas of the femoral condyles. The use of this technique on the patella, trochlea, and posterior femoral condyles requires open arthrotomy. Additionally, the treatment of defects greater than 12 mm in diameter also require open arthrotomy to ensure perpendicular graft placement.

Advantages of this technique include preservation of hyaline cartilage, relative low cost, and the ability to be performed arthroscopically as an outpatient procedure in a single stage. Postoperative limitations are less restrictive than the other techniques, requiring only a few weeks of protected weight bearing. The results appear to be at least comparable with MST in the short term.

There are several disadvantages to this technique. First of all, it is technically demanding and requires advanced skill to be performed correctly. Obtaining perpendicular graft harvesting and placement is critical, along with reproducing an accurate radius of curvature. Very large lesions exceed the amount of available graft tissue. Generally, defects< 2.5 cm in area are most appropriate for this technique to not surpass the capacity of graft sources. There are limitations based on lesion location, and this method is not recommended for patellar or tibial lesions. Far posterior femoral defects are difficult to access with current instrumentation. The harvesting of autologous graft tissue necessitates a certain degree of graft site morbidity, although very little is known about this. There seems to be negligible graft site morbidity in the short term, but no long-term studies have been done to reveal potential long-term adverse effects. Simonian et al measured contact pressures throughout the knee and determined that there are no true nonweight-bearing areas of the knee, although the supero-lateral aspect of the lateral femoral condyle above the sulcus terminalis and the supero-medial area of the intercondylar notch showed significantly less contact pressure [80]. Other issues with this technique include cartilage thickness mismatch, graft impaction forces, fibocartilaginous fill around the plugs, and potential remaining stress riser production between the two nonbinding cartilage surfaces. Multiple small plugs allow better radius curvature matching versus a single large plug, but each graft will be weaker in torsional strength, and more susceptible to delamination during impaction. Grafts > 8 mm in diameter raise concerns with harvest morbidity and radius mismatch. Our preferred plug size is 6 mm.

Overall, autologous osteochondral grafting provides a very reliable means of reestablishing viable hyaline cartilage to a lesion site Fig. 3, Fig. 4.

View Large Image Download to PowerPoint View Large Image Download to PowerPoint View Large Image Download to PowerPoint

Fig. 3. (A) Schematic of graft/inserter placement. (B) Schematic of final impaction of graft. (C) Schematic of final graft placement. (From Levy AS. Osteochondral autograft for the treatment of focal cartilage lesions. In: Cole BJ, editor. In: Cole BJ. Operative techniques in orthopaedics, vol. 11, no. 2. Philadelphia: WB Saunders; 2001. p. 110–111.

View Large Image Download to PowerPoint View Large Image Download to PowerPoint View Large Image Download to PowerPoint View Large Image Download to PowerPoint View Large Image Download to PowerPoint

Fig. 4. (A) Osteochondral grafting, chondral lesion. (B) Osteochondral grafting, roaming recipient site. (C) Placing graft in recipient site. (D) Three osteochondral grafts in place. (E) Second look, 1-year later.

Allograft osteochondral grafting 

The use of osteochondral allografts for tumor treatment dates back to the 1950s. In 1959, Campbell et al reported using fresh osteochondral autografts with little or no destruction occurring over the first year [72]. When fresh and frozen allografts were used, however, they became fibrillated and degenerated within a few months.

The important issues with osteochondral allografting include chondrocyte viability, graft incorporation and remodeling, host immune reaction, and potential for disease transmission. Historically, osteochondral allografts were implanted ”fresh” within 1 to 2 days of harvesting. This presents certain logistical difficulties as the patient and surgeon have to be “on call” if a donor should become available. This also does not allow sufficient time to process and screen specimens adequately for disease transmission risk. As time elapses after harvest, chondrocyte survivability diminishes progressively. It is not known what critical percentage of viable chondrocytes is necessary for optimal long-term durability of implanted cartilage, but it is intuitive that lack of active chondrocytes will lead to eventual cartilage degradation. Preservation methods such as fresh freezing, freeze-drying, and cryopreservation have been studied to prolong the shelf life of specimens, but all have inferior results when compared with fresh grafts [81], [82]. Chondrocyte survival has been improving, however, with newer two-stage cryopreservation techniques. Cryoprotectant use has subsequently produced 85% to 95% chondrocyte survival rates with maintenance of chondrocyte phenotype [83].

Typically, osteochondral tissue has been considered “immunologically privileged” because many believe that it does not generate a host-immune response. Certainly, chondrocyte antigens have been shown to be shielded from the humoral circulation by cartilage matrix, allowing them to avoid recognition by the host-immune system [84]. Though overt tissue rejection does not occur as it would in soft tissue organ transplantation, such as the heart or liver, which requires immunosuppressive therapy, more subtle tissue incompatibility processes may still be occurring. There is evidence to suggest that the bone marrow elements are capable of generating a cell-mediated host response that may impair graft incorporation and remodeling [85].

A major factor associated with the poor results of osteochondral allografts appears to be related to the normal remodeling of the cancellous bone. The deterioration of the articular cartilage in allografts appears to occur as a result of subchondral collapse produced by bone turnover in the metaphysis [12], [86]. Thus, the fate of the transplanted articular cartilage in allografts is related to the fate of the subchondral bone. It appears that for an allograft cartilage transplant to be successful, the allograft must heal to the surrounding bone incompletely so that extensive revascularization and resorption does not result in cartilage collapse.

To date, no studies exist that address the use of modern osteochondral grafting techniques for allograft. Theoretically, if successful, this would improve graft thickness and radius of curvature mismatch concerns in autograft techniques.

We feel that,with all the limitations currently facing osteochondral allografting, it should not be considered an initial treatment of choice. It is a major reconstructive procedure that precludes all other repair techniques. If the graft fails, the fall-back procedure becomes prosthetic arthroplasty. In keeping with our philosophy of not “burning bridges” whenever possible, we would consider osteochondral grafting a salvage procedure and not for initial treatment.

Autologous chondrocyte reimplantation 

Cartilage regeneration implies the transplantation of functional cells or precursor cells into the defect. Regenerative techniques include autogenous perichondral transplant, autogenous periosteal transplant, and autogenous chondrocyte transplant.

Rib perichondrium has been utilized in an attempt to fill osteochondral lesions in young patients. Its use has not been documented in isolated chondral lesions. Early success in osteochondral lesions did not persist because of the high propensity for perichondrium to calcify. Its use is not recommended in chondral lesions.

Periosteum is plentiful throughout the body, although its thickness diminishes with age. Several studies have demonstrated that the cambium layer of periosteum can differentiate into chondrocytes and form hyaline cartilage [7]. When sutured to the base of a lesion, a hyaline-like repair tissue was formed with similar histiologic characteristics (87% type 2 collagen) and biomechanical characteristics as normal hyaline cartilage. The use of continuous passive motion (CPM) has been found to be advantageous in increasing the success rate of periosteal transplantation in animals [87]. Preliminary data has demonstrated the efficacy of this technique in reducing pain in lesions of the patella and the femoral condyle [88]. Controversy exists as to whether the periosteum is best placed with the cambium layer facing into the joint [87], [88], [89] or into the lesion [20], [21].

The advantages of periosteal transplant are that the supply is virtually unlimited and that minimal instrumentation and commercial supplies (cell culturing) are required. Unfortunately, the promising early results of periosteal transplant to the femur appear to diminish with time [90]. Patient morbidity is fairly high, as this requires a formal arthrotomy and prolonged CPM. Additional concerns exist as to the fact that the cellularity of the periosteal harvest appears to be closely related to the experience of the harvester [90].

In order to augment the periosteal transplantation, in 1989, Grande et al reported placing cultured autogenous chondrocytes under periosteum sutured directly to the surrounding hyaline cartilage. In rabbit defects, this was found to produce greater filling of the defects than in controls (periosteum alone sutured to cartilage). Histologically. this repair tissue was felt to be significantly more hyaline-like than the controls. In 1995, Brittberg et al reported on the results of this technique in 23 human subjects [62]. They reported eight good (occasional pain and swelling), six excellent, and two fair, with poor results on femoral lesions 1 year post-transplant. Patellar lesion treatment did not perform as well—yielding two good results, three fair, and two poor. Arthroscopic second looks revealed that there was soft repair tissue in the defects without incorporation into the surrounding hyaline cartilage. This conflicted with the data from previous studies that showed early degeneration at 12 weeks post-transplant [91]. These improved results were attributed to changes in postoperative regimen.

Recently, long-term follow-up has been reported by Peterson et al with a group of 61 patients who have been treated with autologous chondrocyte implantation (ACI) for femoral and patellar lesions [76]. These researchers found that 82% of patients had a good/excellent result after 2 years, and 84% had a good/excellent result after 5 to 11 years. They concluded that most failures occurred within the first two years of surgery, and that if patients had good function at 2 years, they would most likely continue to do so. This suggests that successful ACI results in durable regenerated tissue. In 8 of 12 patients, biopsy of regenerated tissue revealed hyaline-like matrix composed primarily of type 2 collagen, well anchored to the subchondral bone. The remaining biopsies were predominantly fibrous in nature.

Recent data from the Cartilage Repair Registry reveals that 81.8% of patients who had ACI for femoral lesions were significantly improved after 6 years [92]. Medial femoral condyle lesions were most numerous, and 70% of these patients experienced improvement. Lateral femoral condyle and trochlear lesions tended to do better but presented relatively infrequently. The function of improved patients was typically assessed as “good” because there were no “excellent” ratings. The majority of these patients, however, were rated as poor prior to implantation, and ACI was often used as a salvage treatment in very severe cases. Most frequently, reported adverse events include hypertrophic repair tissue and intra-articular adhesions.

Enthusiasm for this technique in the United States has been hampered by theoretic and economic concerns. This staged procedure is invasive and requires a formal arthrotomy. It is also expensive and requires prolonged postoperative limitations. Expectations in this patient group tend to be more modest and directed toward returning patients to activities of daily living as opposed to sports. This may be because this technique is often employed in patients with severe conditions involving very extensive disease and history of previous failed surgery. Recent follow-up provides promise for this technique, however, in properly selected patients. Use of this technique in younger ACL-injury patients with large lesions in knees that are otherwise relatively intact may produce more good-to-excellent results and make returning to sports a realistic expectation (Fig. 5).

View Large Image Download to PowerPoint View Large Image Download to PowerPoint View Large Image Download to PowerPoint View Large Image Download to PowerPoint

Fig. 5. (A) Delamination of chondral lesion. (B) Autologous chondrocyte reimplantation. (C,D) One-year follow-up—hyaline cartilage ingrowth.

Cartilage stimulation via growth factors 

An exciting area of cartilage repair and regeneration involves the use of growth factors. These polypeptides attach to chondrocyte cell surface receptors (integrins) and influence matrix production, proliferation, and migration and replication [7]. The two most promising factors at present are insulin-like growth factor I and transforming growth factor beta. These have been shown to enhance matrix synthesis in animal studies. Unfortunately, these growth factors are not specific for chondrocytes and can affect an entire joint adversely. Further research is needed into delivery, regulation, and modulation of these factors before they are utilized in human joints.

Authors' preferred treatment 

Symptomatic lesions require treatment. The presence of a symptomatic chondral lesion may or may not be known prior to ACL reconstructive surgery. When physical findings of pain, joint line tenderness, and catching or locking are present, it is seldom clear whether meniscal or chondral pathology or both are responsible. As previously stated, MR imaging is of limited benefit as chondral lesions are identified by MR imaging in only a minority of cases. MR imaging is fairly sensitive in detecting meniscal tears, but the mere presence of a meniscus tear on a preoperative scan does not preclude a coexistent chondral lesion that may be causing either a portion or all of the symptoms. Certainly, a normal appearing MR image in the presence of mechanical symptoms is even less helpful. A symptomatic lesion can be considered one whose existence is apparent on arthroscopic viewing in a patient with the appropriate preoperative signs and symptoms. Treatment options in this context include MST, AOG, or ACI.

Alternatively, chondral lesions may be found in patients with no preoperative mechanical symptoms. We believe that asymptomatic lesions discovered incidentally at the time of arthroscopy for ACL reconstruction should be treated with at least a marrow-stimulating technique. There is very little disadvantage to this approach in that it does not “burn any bridges.” Minimal additional morbidity is incurred, and any of the more extensive chondral resurfacing techniques can still be performed at a later date if the need arises with no adverse effect. ACL injury patients are typically very active and place high demands on their knees. Even though researchers have not been able to prove that cartilage repair or regeneration has any effect on the development of degenerative change in the knee, it makes sense that some attempt should be made to address the deficient articular surface in this setting, particularly if the additional procedure adds very little in the way of morbidity. If lesions are on the central and anterior third of the femoral condyles, the surgeon may be justified in performing an osteochondral graft. The harvesting of up to three plugs from the lateral notch can be performed as part of the notchplasty.

Postoperatively, the ACL protocol needs to be modified. Early active range of motion, exercise bicycle, and isometric quadriceps exercises are initiated while the limb is kept nonweight-bearing for 3 weeks. From 3 to 6 weeks, partial weight bearing is initiated. After 6 weeks, the patient is advanced to full weight bearing, and strengthening exercises are instituted. Sports-specific rehab is initiated at 10 to 12 weeks with full return to activities based on a functional assessment.

MST for a chondral lesion should not be termed a treatment failure until symptoms continue for at least 6 months postop. If pain and/or swelling persist and function is diminished, consideration can be given to a salvage procedure. In salvage situations, it is important to allow the subchondral plate to recover from the initial surgery. Once a patient has failed a less-aggressive treatment option, our preference is to perform an ACI technique for lesions of the femoral condyle and trochlea. It is important to begin preparing the patient for this procedure and its implications prior to harvesting the chondrocytes for culture. If this is unachievable for socioeconomic reasons, consideration is given toward osteochondral grafting via a formal open arthrotomy.

ACL-dependent scenarios 

Summary 

The treatment of articular cartilage lesions remains one of the great challenges facing orthopedic surgeons today. The technique of chondrocyte transplantation has opened the door for the application of biologic solutions to difficult problems. These techniques will prove the keystone of further advances into biologic joint repair and replacement. Enthusiasm, however, must be tempered by the numerous gaps in cartilage science and the overwhelming need for further long-term data to demonstrate the efficacy of these techniques in thwarting the presumed eventual progression of these lesions toward osteoarthritis. The status of the articular cartilage is of paramount importance in ACL decision-making. Every effort must be made to protect the existing hyaline articular cartilage during ACL reconstruction. Though current cartilage repair techniques are in their infancy, they remain stepping-stones to future developments. It is hoped that we will one day be able to regenerate normal hyaline cartilage without great morbidity. At present, the ACL surgeon must accept techniques that diminish symptoms and do not burn bridges to future advances. The orthopedic surgeon must increase his knowledge of the basic science of articular cartilage in order to best choose from the various cartilage treatments that evolve.